Dictionary proposal - pydima
|From:||Peter Clark <peter-clark@...>|
|Date:||Friday, June 14, 2002, 17:35|
As I made mention several months ago, I have been looking for a dictionary
program so that I need only enter data once and get a full bi-directional
dictionary spat out, rather than having to duplicate the same information as
I would have to if I went the pen-and-paper (or plain text file) approach. So
far, the only such program that seems advanced enough is SIL's Shoebox, which
costs $45, is closed-source, and runs only in Windows and Mac (and not
Linux). The other option is Boudewijn Rempt's kura program, which is still a
work in progress and which I also have been unable to run. It also has
different goals that what I have in mind; kura seems to be striving to be a
ethnographic database, which is very cool, but all I want at this point is a
dictionary that can be displayed in HTML or LaTeX (possibly RTF, if someone
knows that format).
I did some more research, and found that SIL uses MDF (Multi-Dictionary
Formatter) to take formatted data and convert it into a dictionary. What's
nice, however, it that the manual lists all the tags and usages, so it would
be a relatively easy matter to replicate its functionality.
Thus, I would like to make a Python program (under the GPL) that would be
able to read a text file of data and produce a bi-directional dictionary with
a minimum of fuss. I propose "pydima" (for PYthon DIctionary MAker) as the
name of this project and welcome any help that I can get. It would be nice if
someday there were a GUI interface with it, but at the moment, I can handle
working with a text data file; it's the dictionary that I am aiming for.
The goals of this project are to produce a program that can create
nice-looking, well-formatted, consistant dictionaries with a minimum of
requirements (no MySQL or other database software). It should be able to
expand with the addition of tags (since conlangers are a creative bunch) and
I am currently writing up a more in-depth guide that outlines how things
would work ; if you are interested, please write to me for more information.
I will be upfront: I am not a coding genius. I used to do some, but my skills
(what little there were) have lapsed. That is why I am approaching the group,
because while I believe I can do this, I know that it will go faster if there
are two or three others who would be willing to chip in.
Thank you for your attention; we now return you to your regularly scheduled
 http://www.sil.org/computing/shoebox/MDF_2000.pdf (Warning, big file:
2.2MB) The tag descriptions begin on page 20. For a concise listing of all
the tags, see http://www.sil.org/computing/shoebox/MDF_AppendixA.html
 I chose Python because it is widely known, cross-platform, less of a
maintenance headache than PERL, and it's the only language I am really
comfortable with. Sorry.